Front page | perl.module-authors |
Postings from November 2011
Re: The CPAN Covenant
From: Lincoln A Baxter
November 29, 2011 04:20
Re: The CPAN Covenant
Message ID: 1322569222.2299.82.camel@lws
On Tue, 2011-11-29 at 08:26 +0200, Gabor Szabo wrote:
> On Tue, Nov 29, 2011 at 7:17 AM, <firstname.lastname@example.org> wrote:
> > I've mostly ignored all this but I will step in briefly.
> > I am the author of a few modules on CPAN. None of them are major.
> > Other stuff of mine is in the alpha, proof-of-concept stage as a set.
> > I have various sysadmin tools for DNS which I haven't finished ...
> Let me jump in on this as I think these are very good examples for
> why are there many modules that *seem* not to be useful or
> maintained. These might bring to confusion, especially for the 99% who
> are not in the echo chamber.
> I have a similar list though I think, unlike those of dhudes,
> most of my stuff is totally useless.
> Let me give some use cases: (or maybe useless cases) of
> *my* CPAN modules:
> 1) Some are totally useless. The reason I don't remove them is
> a) someone, somewhere might be using it
> b) be able to learn from it (e.g from the mistakes)
> c) I have emotional connection to them (my first one)
> I'd be happy to mark the modules accordingly if there was an easy way
> to do that but I don't want to upload a new version for it.
> 2) I have some modules that might be useful but I lost interest in them.
> It would be nice if there was an easy way to mark them (again, without
> uploading a new version) that they are (IMHO) somewhat useful and they
> are up for adoption.
> 3) I have a few modules that would need some serious work. Currently I don't
> have time but I hope in the future I will. I don't want to give away
> maintenance but I might be open to patches. I'd like to be able to
> easily express this without uploading a new version.
> 4) There is I think only one distribution that is really useful and there
> are multiple authors and multiple co-maintainers for that.
> So that's OK I think.
> So if there was an easy way to mark these things I'd probably put in
> some time to do that. Then this could be displayed on the
> interfaces CPAN has.
> (I don't mind if this is in PAUSE or MetaCPAN)
> Some other - active - people might also be ready to mark their modules.
This is a good idea. Then users of CPAN search could filter their
search results, using this information. I too have a module that I
would mark as unmaintainable by me (Openview::Message), because I no
longer have access to the underlying software. I'd love to so mark that
one as unsupported. I have not updated it since Jan of 2004, and would
not want to update it just to add something to the read me that says
My other module (Sys::SigAction) has been positively reviewed, and even
included in a Debian package. (They asked me to change the license to do
that, and I did). While I no longer get paid to program in perl, I
still love the language, and this module is one I do make time to work
on when there are problems. Recently I developed a new capability as a
result of one of the CPAN reviewers comments.
> This does not solve the problem for the modules of authors who have
> not been around for a long time but this might be a good step
> in improving the situation and its perception.
Yes, the reviewer whose comments stimulated my improvements to
Sys::SigAction is no longer reachable, though he has a PAUSE id. The
ability to mark a module as actively supported (and perhaps the
need/ability to update that marking periodically), could certainly be
used as a filter for someone searching CPAN. (For instance one could
look for modules marked as actively supported within some date...)
Or, here is another idea (which I have not seen broached), that could be
completely automated fairly easily: Mark those modules as unsupported
and open to co-maintainer assignment, where all of the maintainers'
PAUSE id email addresses BOUNCE mail messages, . This could be used to
grease the assignment of co-maintainer status, and in CPAN search
filters. Obviously, if the PAUSE id is not being maintained by an
author, the author has probably lost interest and is difficult to reach
without significant effort, if at all.